Diagnostic Imaging
June 1999

The Digital Department
Point/Counterpoint: Why radiologists should maintain control of PACS

CIOs don’t understand the extent to which PACS require a redesign of the operation of the radiology department


Pogo must have been a radiologist. His infamous words—“We have met the enemy and he is us”—can be applied to all of medicine, but seem especially apropos to our specialty.

While we lament the depredations of our clinical colleagues, radiologists by their inaction have squandered diagnostic opportunities. We now have a tool at our disposal—PACS—that if used properly may allow us to regain a larger role in patient care. It has initiated a new turf war, however, with an unexpected member of the healthcare team: the information technology (IT) specialist.

IT departments are becoming larger and more influential as hospitals increasingly depend on computer systems for financial analysis, utilization management, and patient-care data retrieval. As hospitals begin to move toward an electronic medical record (EMR), it is natural that questions arise about the incorporation of radiologic and other medical images into the EMR.

At the Baltimore VA Medical Center, many of the people who make decisions about the PACS are members of the IT department or the hospital administration. The IT personnel often express the same genuine but misguided opinion that PACS is just another information system. They also believe that radiology images are just big data sets made up of bits and bytes, like any other part of the EMR.

Vendors of hospital information systems and radiology information systems are responding to this situation by incorporating image management capabilities and radiology workstations into their systems, allowing distribution of radiology images throughout the healthcare enterprise.

Despite the decreasing role of radiologists in the purchase, implementation, and management of PACS, we strongly believe that responsibility and authority for the PACS must remain the domain of the radiology department. Our concern stems from the need for an appreciation of what radiology can accomplish.

Chief information officers (CIOs) don’t understand the extent to which PACS require a redesign of the operation of the radiology department, nor do they comprehend how important optimizing work flow is to the success of an imaging department. As Dr. Richard Morin, a professor of radiology at the Mayo Clinic Jacksonville (FL), has said: “The proper and efficient function of PACS is directly linked to the department work flow. Only the radiology department can properly design, configure, and monitor PACS’ performance.”

Without a detailed understanding of the operation of an imaging department, a CIO who believes PACS to be just another information system will inevitably create a system that squanders human and equipment resources. This PACS will, at best, result in a system with an inefficient work-flow design; at worst, it will cause an imperfect but functional film system to be replaced by a useless white elephant.

By virtue of our training, radiologists are experts in the maintenance and assessment of image quality. IT staff must realize that good patient care requires a complete understanding of image quality control. Upside-down and mislabeled radiology images are good for a laugh when seen in a movie or on TV, but they are downright scary when seen on hospital information systems at national IT conferences.

Although we at the Baltimore VA have the good fortune to have excellent staff, the expertise of IT staff and their level of responsiveness vary greatly among facilities. Many IT departments do not meet the requirement for 24-hour, seven-day-a-week coverage and uptime. While a response time of four or more hours may be acceptable for a hospital information system, it is totally unacceptable for a filmless radiology department and hospital. IT staff must understand that uptime for the system must approach 100%, including maintenance and updates. This requires a level of reliability and redundancy that exceeds current requirements for hospital information systems.

Radiologists are able to solve problems associated with various components of the PACS in a way that would not be possible for IT personnel. Two simultaneous calls about workstation malfunctions, one from a panicky medical student preparing for daily rounds and the other from an orthopedic surgeon in the operating room, might be prioritized differently by a radiologist and an IT staff member. Additionally, many physicians distrust IT personnel, who are notorious for making decisions that affect healthcare delivery without first consulting physicians.

Radiology departments that have made or are making the transition to filmless operation realize that PACS represents much more than an information system and is actually a “medical device” that enhances our ability to diagnose and manage patients. Computer workstations are becoming examination tools for radiologists just as stethoscopes are for medical internists. During the next 20 years a variety of new diagnostic tools will be available to radiologists for image optimization and computer-assisted diagnosis. The selection of the tools must remain the prerogative of radiologists.

We applaud the trends toward an EMR, improved patient data collection and analysis, and improved hospital computer and communications infrastructure, all of which benefit diagnostic imaging and patient care. We believe that radiology and IT departments should work together to ensure the full integration of radiology information and PAC systems and imaging modalities.

The RSNA and the Healthcare Information and Management Systems Society have joined in a collaborative project aimed at promoting connectivity between modalities and information systems in the radiology and nuclear medicine departments and enterprise healthcare information systems. This follows a similar initiative in 1992 between the American College of Radiology and the National Electrical Manufacturers Association, in the formation of DICOM connectivity standards. These types of collaborative efforts represent the ideal synergy between two diverse groups of healthcare professionals, each with a vested interest in the overall success of PACS. Similar efforts between IT staff and radiology departments are essential to the success of a PACS. But radiology should not relinquish management authority and responsibility in the interest of these collaborations. Control of PACS must continue within the domain of radiology.

DR. SIEGEL is chief of radiology and DR. REINER is director of radiology research at the VA Medical Center in Baltimore.

You’re wrong! Why IS staff should assert its control

Images are merely big data sets, similar to any other patient data set


Hospital chief information officers have been arguing for years with physician groups over who should manage application systems running in their departments. The battle has lately shifted to PACS as it becomes a cost-effectiveness tool for radiologists. While there might have been some discussion about this in the past, there is today no question that PACS should be managed by CIOs and information systems staff. A quick look at the history of computing in hospitals makes clear why this is necessary.

The healthcare industry has long been known for being among the least innovative of American business sectors in its use of information technology. Hospitals embraced IT only when they had to, not because they wanted to. The first use of computers in hospitals 30 years ago reflected the industry’s response to Medicare and Medicaid billing requirements. It is easy to forget that before the government became involved, hospitals frequently billed a global, all-inclusive daily rate. Legislation and cost-based reimbursement established the incentive to track every item. It then made more sense to computerize billing than to employ an army of clerks posting charges.

Because the first hospital computer applications were financial, the data processing department usually reported to hospital accountants. Clinical information systems were virtually nonexistent. If clinicians wanted a computer in their department, they didn’t usually get very far with the DP department.

Departmental information systems got their start because data processing was so unresponsive to clinicians’ needs. DP was focused on finance, and companies like IBM developed software to perform financial functions. Many users, not just those in healthcare, saw the need for computers that could perform scientific as well as business functions. The development of the first minicomputers offered a lower cost platform with scientific programming languages. Minicomputers lent themselves well to the clinical environment.

The late 1970s and early ’80s saw the development of the first hospital applications that bridged administrative and clinical information systems. The first order entry and results reporting systems, sometimes called hospital information systems, offered the promise of speeding ancillary orders and results while improving efficiency. But the departmental systems now had to talk with the HIS. This meant that complex interfaces had to be developed, sometimes in different proprietary programming languages. Interfaces became the nightmare of the DP department. This problem was ultimately resolved when software vendors adopted the Health Level-7 standard, but the process took many years.

About the same time, the role of the data processing director was changing. More “customers” came not from finance but from clinical operations. The DP director had to focus on all of the functions in the hospital. The title of data processing director gave way to information services director and later to chief information officer. By the late 1980s the CIO had become the person within the hospital hierarchy most involved with how the entire organization managed information and technology. The CIO was expected not only to manage the technology infrastructure, but to stay abreast of changes in technology.

PACS was born in the radiology department because of radiologists’ adoption of new digital modalities. Many CIOs were excited about what was happening in radiology because it was a clinical area where computer technology directly impacted patient care. But like the early lab systems, there was no interest in directly linking a CT or MR system with hospital computer systems—because there was no need. Furthermore, PACS technology was extremely expensive and most CIOs recognized that a dramatic decrease in costs was essential if the technology was to become cost-effective.

The past five years have seen a major change in the computing environment of the healthcare CIO. Today’s CIO is most likely in charge not just of a hospital’s computers but of an entire healthcare delivery system that may include several hospitals, outpatient centers, and sub-acute and long-term care facilities. The focus of the CIO has moved from the computer room to the network, which has become the center of the computing universe. There is more computing power sitting on the desktop than ever resided in the computer room. The most daunting challenge facing the CIO is how to effectively manage applications when there is limited control over the desktop.

Because of advances in programming languages and the development of database tools, the once-elusive electronic medical record is moving closer to reality. Web technology offers the possibility of accessing patient information from the Internet. Higher speed networks and the next-generation Internet have the potential to bring multiple pieces of information together to support the clinician.

These same advances have made PACS a viable product. Storage technology advances and faster networks have overcome some of the earlier limitations of PACS. Diagnostic workstations can now run on Wintel PCs, eliminating the need for expensive RISC platforms. The acceptance of the DICOM standard has made it easier and more cost-effective to link the PACS components together. PACS can now be realistically considered for smaller hospitals and outpatient centers.

So why does PACS need to be managed by the CIO and information systems department? Because it is nothing more than another information system. Digital images of x-rays are just that—digital images. They are big data sets that need to be managed in the same manner as any other patient data set, with an emphasis on security and confidentiality. The fact that it is possible to integrate the images with the radiologist’s report and all the other clinical information makes it imperative that the CIO be in charge. In fact, this is the fundamental role of the healthcare CIO, who has the express responsibility for the EMR. PACS is just one component of this.

The CIO and IS department are responsible for maintaining the network infrastructure and ensuring that it works well. PACS will achieve its maximum patient-care benefit when the images are available from the same workstation as other clinical information. Radiologists are not equipped to become network managers. Even when a separate subnetwork is created for linking the diagnostic workstations, it is more appropriate to have the same network technicians supporting both.

The CIO has fiduciary responsibility for ensuring that electronic patient information is secured, handled confidentially, and safeguarded appropriately. CIOs must have elaborate policies and procedures in place to ensure that data are protected. They must have plans for recovering information in the event of a disaster. All of this must include the PACS as well. Radiology department management of PACS would result in a duplication of functions and the possibility that some vital function might not be performed at all.

PACS, like so many other departmental application systems, was developed by the radiology department, and 10 years ago, it could have been argued that there might be merit in allowing that department to manage it. In today’s complex hospital environment, however, PACS must be managed by the person who has responsibility for integrating it into the rest of the electronic patient record. That person is the chief information officer.

Editor’s note: This point/counterpoint does not end here. Plan to attend the PACS conference, “Healthcare I2: Bringing Imaging & Information Together,” which will include a general session on “Who’s in Charge of PACS?” featuring Drs. Siegel and Reiner and Mr. Bowers.

MR. BOWERS is chief information officer for American Radiology Services and was formerly CIO of the University of Maryland Medical System.

Back to Homepage | New Search

Copyright © 1999 Miller Freeman, Inc., a United News & Media company.